<p>Agro-industrial by-products are increasingly regarded as suitable matrices for recovering bioactive compounds. In this context, solvent selection is critical for both analytical characterization and process design. This study aimed to identify a solvent system for extracting polyphenols from yerba mate (<i>Ilex paraguariensis</i>) leaf residue and to assess the effect of successive extraction cycles on overall recovery. Antioxidant capacity and sample-preparation greenness were also evaluated for exhaustive extracts. An augmented simplex-lattice mixture design was applied using water, ethanol, and acetone. Data were fitted to a full cubic model describing polyphenol content as a function of solvent composition. The model revealed synergistic interactions between water and the organic solvents, identifying two binary mixtures that maximized extraction: water–acetone 45.2:54.8 (w/w) and water–ethanol 47.3:52.7 (w/w). Successive extraction showed that water–organic mixtures required four cycles to reduce polyphenol content below the limit of quantification, whereas water alone required one additional cycle. After exhaustive extraction, HPLC-DAD quantification of seven target polyphenols yielded totals of 107.3, 114.4, and 119.2&#xa0;mg&#xa0;g⁻¹ (dry weight) for water, water–ethanol, and water–acetone, respectively. Mono-caffeoylquinic acids were similarly recovered across solvents, whereas water–organic mixtures enhanced the extraction of less polar polyphenols (di-caffeoylquinic acids and rutin). Antioxidant capacity assays showed no significant differences among exhaustive extracts prepared with different solvents, while AGREEprep indicated a greener sample preparation for water than for aqueous–organic mixtures. Overall, this study provides quantitative criteria for selecting solvents for applications prioritizing recovery or compositional profiling in yerba mate residues.</p>